Pulse transformer for Half Bridge

Hello Friends,
Any one have a good pulse transformer based half bridge driver design?
Please give reference about turns of primary and secondary, Mosfet protection.

Hi,


You surely have already done a search here in the forum and with google. There are a lot of discussions, schematics and application notes around.

What exactely are your needs and why do you know you need puse transformers and give no other information like circuit, voltage, current, frequency, duty cycle... and all your other ideas.

Personally i try to avoid pulse transformers with MOSFET halfbridge, but i use them with controlling high voltage SCRs..

Klaus

I am working an ultrasonic power circuit which must have 50% Duty cycle and power will be 50 watt.
For this i have worked with Irs2110 but not succeed, Here my microcontroller stops working after some time, so i want to go for pulse transformer based circuit.

I know of schemes where the pulse transformer is simply
some distance of paralel PCB traces.

There exist, now, gate driving isolator products like (I
think) from Avago. If you are having isolation problems
this might be a way to go. But you want to look at that
and determine whether the destructive path involves the
IR2110 control inputs per se, or perhaps relates more to
poor ground integrity and injecting spurious signals to
the uC such that it just gets lost.

Pulse transformer drive is actually a very rugged way of driving and isolating Mosfets. However, it is by no means simple, and has several gotchas.

There was an excellent app note written by Unitrode, which described in great detail how to actually implement a gate transformer drive. When Texas Instruments purchased Unitrode, they kept most of the appnotes, and it may still be there. The last time I actually read it was about 2 years ago.

Hello Friends,
Any one have a good pulse transformer based half bridge driver design?
Please give reference about turns of primary and secondary, Mosfet protection.

Click to expand...

Hi deepeshmishra
I can guide you about this but before that , you must deliver your exact information . for instance :
Frequency of operation . mosfets or IGBTs which you've used in your design . and tell about specifications of your PWM driver .


Best Wishes
Goldsmith

goldsmith said:

Hi deepeshmishra
I can guide you about this but before that , you must deliver your exact information . for instance :
Frequency of operation . mosfets or IGBTs which you've used in your design . and tell about specifications of your PWM driver .

Click to expand...

Thanks For taking interest.
My frequency is 30Khz to 36Khz, I am using Mosfet Irf840 and I am generating frequency from microcontroller in half bridge mode .

HI,

just out of curiosity i googled for "pulse transformer mosfet"
(in fact i was interested in the solution to avoid saturation)

on the first page of of the google results there are at least 5 application notes describing the whole circuit, problems and solutions very detailed.

I openend one of the application notes and got my wanted answer within about 10 seconds.

@deepeshmishra: now you wait more than 360.000 seconds for an answer

Why do you want a pulse transformer to drive a half brige?
(usully there is no need to isolate the driver)

What is your circuit?
I´ve done a google search with "ultrasonic piezo driver circuit". There are a lot of schematics. Which one do you use?

Usually pizos need high voltage. Do you want the half bridge output to directely drive your piezos?
You can also use a 24V halfbridge and use a transformer to get high voltage.

******
So please do a research and decide which way to go.
Then go through the application notes and tell us where exactely you need help. (give pictures, links, technical data, ....)

I am sure you get your answers within some hours.


Klaus

if you can't find a cheap pulse transformer, for that frequency range you should find most 4mH common mode chokes to be sufficient. as far as driving the transformer, most of the issues people run into is trying to get 100ns or less rise times. its a 36Khz half bridge.. switching time is not at all critical.. not to mention that you have partial zero current switching anyways so you could probably get away with just about anything.

you should be able to get away with fairly small surface mount transistors to drive the transformer(s)

Thanks For taking interest.
My frequency is 30Khz to 36Khz, I am using Mosfet Irf840 and I am generating frequency from microcontroller in half bridge mode .

Click to expand...

Hi again
First you must calculate the maximum required Gate current , i suppose you know how , isn't it ?
After , that , you can follow this instruction in order to design a suitable pulse transformer , read the topic which comes in below , from first page up to the latest page you'll catch a lot of things , i'll bet !

https://www.edaboard.com/threads/229615/


Then you should follow this circuit :



Best Wishes
Goldsmith

I would make one simple addition to Goldsmith's excellent drive circuit.

Add a small resistor (10 to 50 ohm) in series with left capacitor.

This will dampen any ringing between the capacitor and the leakage inductance.

Add a small resistor (10 to 50 ohm) in series with left capacitor.

This will dampen any ringing between the capacitor and the leakage inductance.

Click to expand...

Hi schmitt trigger
I've never saw anything like what you said before , and all of my transformer based drivers was without that resistor , but some minutes ago i've tested that and you were right , the waveform looks a lot better than before . may you explain it that how exactly that will affect the circuit please ?
Thanks

Goldsmith

All transformers have leakage inductance, which essentially means uncoupled inductance.

This can be modeled as a series inductor. With the series capacitor it will form a high-Q LC tank that will resonate.
Depending on the actual switching frequency, and the actual LC values, this resonance could become excessive.
The series resistor will help dampen that resonance.

But its value must remain low, otherwise you will limit the actual drive current.

The best value is often based on empirical measurements.

All transformers have leakage inductance, which essentially means uncoupled inductance.

This can be modeled as a series inductor. With the series capacitor it will form a high-Q LC tank that will resonate.
Depending on the actual switching frequency, and the actual LC values, this resonance could become excessive.
The series resistor will help dampen that resonance.

But its value must remain low, otherwise you will limit the actual drive current.

The best value is often based on empirical measurements.

Click to expand...

Hi again
Thanks for the explanations . i must confess that i have never looked in this issue like this . impressive .
Thanks again
Goldsmith

Thanks for replying friends I have done. some resonant tuning needed that will be done.